Transcript If sidelobe is high, RFI looks like ETI signal from star

Funded by Paul
Allen and Nathan
Myhrvold
Developed jointly
by SETI Institute
and U. C. Berkeley
RAL
ATA-3 (now), ATA-32 (2005)
ATA-350 (2008)
Do SETI on 4 different
stars at once. Sidelobes
from each beam are
uncorrelated. RFI will
usually appear in more
than one beam – can be
discarded.
0 dB
Place
Beam on
Star
P4 (N) (dB)
Circle: ATA
Primary Beam
0
20
25
2000
4000
6000
8000
10000
Frequency (MHz)
9. Declination
Dependence
Rejection Ratio
Rato vs. Declination
Rejection
350 Antennas, 1420 MHz
P = -50 db  0.001% chance
that RFI is 3 dB higher in
one beam than all others.
-10
-20
-30
-40
-50
0
-10
-20
1 dB HA = 0
3 dB HA = 0
1 dB HA = 45
3 dB HA = 45
-30
-40
-60
-70
0
0
20
40
60
80
1
2
3
4
N (dB)
Sidelobe level (dB below max)
Gather statistics from beam
simulations.
3 dB HA = 45
-50
-30
-15
0
15
30
45
60
75
Declination (deg)
It is very unlikely that RFI can
masquerade as ETI signal.
-40
-15
0
15
30
45
60
75
Declination (deg)
Sidelobes are fatter and less
complex, but technique is
still fairly effective.
A.11. Conclusions
1 dB
3 dB
0
Rejection Ratio (dB)
-10
P4 (N) (dB)
Prob. of sidelobe level (dB)
4
-30
0
-5
-10
-15
-20
-25
-30
-35
-40
-45
-50
6. Better: Take many
snapshots while
source moves.
0
-60
3
Better at higher frequencies
since sidelobes are smaller.
0
-50
2
350 Antennas, 1420 MHz, Dec = 20°
Probability RFI is N dB higher in one beam
than in all others.
-40
1
N (dB)
Log Sidelobe Probability vs. Log Sidelobe Power
1 dB Bins
3 dB HA = 0
-30
Even with only 2 beams, RFI
is well identified.
If sidelobe is high, RFI looks
like ETI signal from star.
-30
-40
N (dB)
-60
15
1 dB HA = 45
-20
0
Lower is better.
10
1 dB HA = 0
-60
-50
-20
-30
-50
-30
-50 dB
-20
RejectionRatio
Ratovs.
vs.Frequency
Frequency
Rejection
5
-10
4 Beams
Probability RFI is N dB higher in one beam
than in all others.
-20
0
3 Beams
-10
P = -5 db  50% chance
that RFI is 3 dB higher in
one beam than all others.
-40
2 Beams
8. Frequency
Dependence
-10
Rejection
Rejection Ratio
Rato vs. Declination
32 Antennas, 1420 MHz
0
5. Snapshot
Observation
0
3. How often does
this happen?
10. Fewer Antennas:
ATA-32
Probability RFI is N dB higher in one beam than in all others
(rejection ratio)
2. Synthetic Beam
RFI Enters
In the
Sidelobes
7. How does it work
with fewer beams?
Rejection Ratio (dB)
350 x 6m dishes
RF = 0.5 – 11.2
GHz
IF = 4 x 100 MHz
4. Trick: Four
Simultaneous Beams
PM(N) (dB)
1. Allen Telescope
Array (ATA-350)
Allen Telescope Array, SETI Institute, Mountain View, CA
Rejection Ratio (dB)
Gerry Harp
Best at low declination since
source moves faster.
Multibeam correlation
is a great way to
exclude RFI from SETI
observations.
ATA supports up to 4
simultaneous beams,
gives high confidence
that RFI will appear in
more than one beam.
Not shown: It is
important to average
power, not amplitudes,
when averaging
snapshots (please
refer to paper).